Dr. Dariusz Lapucha
Fugro Chance Inc.
Dynamic positioning (DP) has become a byword in the floater industry, enabling drilling in previously unreachable deep-water and ultra-deepwater regions. Today, companies that supply differen-tial global positioning systems (DGPSs) are making system enhancements to improve DGPS accuracy and integrity that will expand their application.
Fugro Chance Inc. has created the Starfix High Performance (Starfix-HP) system to answer DGPS needs. The new system reduces noise for DP vessel positioning, improves navigation for the high resolution and deep seismic markets, improves position data for gravity surveys, and provides real-time tide correction. These features translate to significant cost savings and better system performance for the marine market.
Current marine DGPSs provide meter and sub-meter levels of position accuracy and repeatability. Fugro's Starfix-HP system provides decimeter-level accuracy, which increases positioning precision.
In the offshore environment, the approach to improving accuracy is different from the approach on land. Local reference stations are not available. However, the open seas offer a near ideal environment for GPS signal tracking. The primary reason is neither buildings nor foliage impede signals, so the antenna normally has a clear line of sight to all satellites above the horizon. In addition, onboard GPS receivers experience few cycle-slips.
Fugro has combined existing Starfix DGPS satellite data networks and associated reference stations, high-precision dual-frequency carrier phase measurements, and the offshore environment to provide an advanced system accurate to better than 10 cm horizontally and 20 cm vertically.
Gulf of Mexico static results.
Some offshore applications now require this degree of accuracy, which until now has been impossible to achieve or prohibitively expensive using existing technology.
DGPS is an established technique that provides meter level positioning accuracy in real time. Current DGPS systems typically use single-frequency L1 code pseudoranges differenced between the rover and reference station. The accuracy of single baseline DGPS decreases as the distance to reference stations increases. Degradation occurs because of special decorrelation of the tropospheric, ionospheric, and satellite orbit errors.
Outside the box
A number of techniques have been developed to improve the accuracy of single baseline DGPS. For example, using multiple reference stations in a network configuration has been shown to improve the accuracy of DGPS over the single baseline approach. However, fundamental limitations of GPS pseudorange measurement noise and signal multipath limit the positioning accuracy of DGPS to the meter and submeter level.
Real-time kinematic (RTK) positioning is one technique that can potentially provide centimeter-level positioning accuracy. The Achilles heel of RTK positioning is distance. Accuracy in resolving integer carrier-phase ambiguities diminishes when the user is more than 10 km from a reference station. Even with the aid of a network, the RTK technique requires a distance not greater than several tens of kilometers. Because of its fundamental distance limitations, RTK positioning is not well suited to offshore coverage.
Gulf of Mexico kinematic test results.
Fugro's new decimeter-level accuracy system bridges accuracy and coverage gaps between a small area centimeter-level RTK and a wide area meter-level DGPS. Starfix-HP complements Fugro's other geostationary satellite-based DGPS positioning services and opens up new application areas.
How it works
The new, high-precision system uses dual-frequency observations in a network mode to provide a decimeter-level positioning solution. Unlike standard DGPS systems, which use pseudorange code observations, Starfix-HP exploits the high precision of dual-frequency carrier phase observations to achieve a higher accuracy positioning solution. The difference is that pseudorange code measurement accuracy is at the 3-m level, while carrier phase measurement accuracy is at the 2-mm level. This level of accuracy is determined by the ability of the system to measure the fractional wavelength of each of the observations, which are 300 m and 20 cm respectively at the L1 frequency.
Starfix-HP uses a proprietary method to resolve double difference carrier phase ambiguities exploiting satellite redundancy. After ambiguities are resolved, the system maintains decimeter accuracy positioning as long as at least four satellites maintain a lock. Rising satellite ambiguities are recomputed without deteriorating a positioning solution. Because of the inherent precision of carrier phase measurements, Starfix-HP provides continuous smooth positioning without jumps caused by satellite constellation changes.
Solar activity creates ionospheric disturbances that are a major error source for single-frequency-based GPSs. Ionosphere-induced errors are especially severe in lower and higher latitude areas and along the geomagnetic equator. Eliminating ionospheric influence is one of the Starfix-HP system's key achievements and is a basic condition for long-range high-precision positioning. Starfix-Plus, Starfix-HP's predecessor, eliminated ionospheric errors based on dual-frequency code solution. Starfix-HP goes a step further by using ionospherically corrected carrier phase observations to achieve higher accuracy.
Another key element of the Starfix-HP algorithm is the use of multiple reference stations in a network mode. The reference station dual-frequency carrier-phase corrections are processed in a correction domain to produce virtual base station corrections. These corrections are optimized for a user location. The corrections, combined with rover observations, produce virtually error-free positioning at the decimeter level.
Combining multiple reference-station corrections in correction domain ensures full use of all the satellites observed at the rover. As a result, the positioning solution does not suffer from deterioration of satellite geometry of the single baseline solution as distance from the reference station increases. The multi-reference method minimizes the influence of distance-dependent errors and is intrinsically robust against single-station failures. The method is also flexible because it does not require all stations to observe all stations used in a rover solution.
North Sea results.
The system's positioning rate is limited only by the observation rate of the GPS receiver, but is independent of the reference correction update rate. The user algorithm uses the current GPS observations and extrapolated GPS errors inherent in reference station information to derive the optimal solution using the extrapolation algorithm. Continuous, high-precision positioning is maintained without interruption even when reference data reception is intermittent.
In May 2001, Starfix-HP entered into service in the Gulf of Mexico and the North Sea marine markets. Since then, system capability has been expanded to cover the Middle East, Southeast Asia, South America, Africa, and Australia.
Providing a high-precision service means reliably retrieving reference station data to the network control center and broadcast data to the satellite uplink sites. To that end, Fugro employs frame relay, lease lines, very small aperture terminal (VSAT) circuits, Fugro's own data network, integrated services digital network (ISDN) circuits, dial analog circuits, and the Internet. Fugro also has established a density of reference stations so that a single site failure does not affect the end user's positioning accuracy.
The user setup comprises the integrated GPS receiver, L-band receiver, and external processor board. Position fix computations take place in an external processor. The Starfix-HP system supports several major brands of dual frequency GPS receivers. These receivers provide low-latency, high-rate dual-frequency GPS observations that are processed with L-band data to produce decimeter-level position. Further plans include the full integration of the software at the GPS processor level in conjunction with the receiver manufacturers to eliminate the external processor.
Proving the system
The Starfix-HP system has been tested and is being monitored in real time at a number of locations. The monitoring units use the same setups as the real-time users receiving over-the-air reference data information. A test at a static location in Lafayette, Louisiana, used five GoM stations 350 km to more than 1,000 km from the unit.
The sample results demonstrate subdecimeter horizontal and decimeter vertical accuracies. These results are representative of what can be achieved anywhere in the GoM and correspond closely with the results achieved in the Starfix-HP North Sea test.
Fugro tested the Starfix-HP system previously tested on a survey vessel in the GoM in July 2001 to assess performance in a dynamic mode. The real-time position results of the system were compared with post-processed centimeter-level kinematic trajectory. The "truth" kinematic trajectory was generated using state-of-the art kinematic software in backward and forward runs.
For this test, the distances to the GoM reference stations ranged from 400 km to >1,000 km. The local base station was set up onshore for kinematic control at a distance of 10-35 km from the vessel. Distance from the local kinematic base station to the vessel was within 10-35 km. The results of this test confirm that Starfix-HP accuracy is not dependent on the dynamics of the platform.
The Starfix-HP system was also tested in the North Sea, which has significant tides. The independently measured tide curve provides additional means of verifying the height accuracy of the system. The close correlation of the tidal and Starfix-HP height curves indicate the Starfix-HP system's potential to determine the height of a vessel performing a survey without depending on a known mean sea level and external tide measurements.
A new high-accuracy system provides precision an order of magnitude beyond conventional DGPSs over a wide area. With recent advances in GPS receiver technology, this higher positioning accuracy can be achieved reliably in real time at a high rate and at low latency.
Higher positioning accuracy coupled with near-instantaneous output will enhance not only current applications of DGPS, but will also open new applications primarily in broadly defined high-precision navigation and automatic guidance.
In the marine environment, the primary benefit will be easier realization of stable and accurate vertical reference based on homogeneous Earth referenced frame. The vertical reference currently used in the marine environment is based on mean sea level reference. This reference is cumbersome because of temporal and spatial variations of sea-surface topography that appear due to tides, currents, and storm surges. Starfix-HP resolves common ambiguities in tidal charts and tidal computations. Improved knowledge of vertical reference will have significant impact on the reduction of seismic, gravity, bathymetry, and other sea sensors data. This in turn will lead to better mapping of seabed and sea sub-bottom.
Richard Barker is manager of systems engineering and support with Fugro Chance, Lafayette, Louisiana. He has been involved in marine navigation and positioning systems design and implementation since 1979. He received a BS from the California State Polytechnic University, Pomona, California.
Dr. Dariusz Lapucha is senior geodesist with John E. Chance & Associates, Lafayette, Louisiana. He received a PhD from Warsaw University of Technology, Poland, and M.Sc. from the University of Calgary, Canada, both in Surveying Engineering.
Tony Wood is international marine manager for Fugro Chance in Houston, Texas. He has been involved with marine positioning in both technical and managerial roles since 1979. Tony was elected to the Royal Institute of Navigation in 1986. He was educated at technical colleges in Birmingham and London.